1 . The theory of abrogation and the possibility of revision

According to the Islamic view that the quran QurŸån is the speech of God conveyed to muhammad Muøammad by an angel, there can be no revision of the quran QurŸån by muhammad Muøammad of his own volition. This is made clear in a number of verses:

When our signs (or verses) are recited to them as Evidences, those who look to no meeting with us say, Bring a different quran QurŸån from this, or alter it. Say, It is not for me to alter it of my own accord; I follow only what is revealed to me; if I go against my Lord, I fear the punishment of a mighty day. [10.15/16].

An earlier passage described the punishment more vividly:

If he were to forge against us any statements

we should take him by the right hand

then cut his heart-vein;

not one of you would protect him (from us). [69.44-47].

Yet the other verses indicate that the pagan Meccans brought pressure to bear on muhammad Muøammad to produce 'revelations' more favourable to themselves, presumably by permitting some recognition of the idols as lesser deities.

They almost tempted you from what we revealed to you, so that you invented against us something else; and then they would have taken you as a friend. Had we not made you stand firm (muhammad Muøammad), you had almost inclined towards them a little. Then we would have made you taste the double of life and the double of death, and you would not have found against us any helper. [17.73/5-75/7].

muhammad Muøammad must have believed that these were true revelations, and therefore could not have contemplated deliberately producing any verses and passing them off as revelation.

Nevertheless the quran QurŸån speaks of various ways in which changes come about by the initiative of God. God may cause muhammad Muøammad to forget some verses; but, if he does so, he will reveal other verses in their place.

We shall cause you to recite, and you shall not forget

Except what God wills . . . [87.6f.]

For whatever verse we cancel or cause (the messenger) to

forget we bring a better or the like. [2.106/0]

The following verse probably also refers to this, but it could also refer to the forgetting of matters other than revelations:

.... . and remember your Lord when you forget, and say:

Perhaps my Lord will guide me to something nearer the

 truth (rashad) than this. [18.24/3]

There are also verses which speak of God deleting or otherwise removing and changing certain passages.

God will delete or confirm what he will; and with him is the 'mother' of the Book. [13.39]

When we substitute one verse for another-and God knows best what he sends down-they say, You (muhammad Muøammad) are simply an inventor; nay most of them do not know. [16.101/3]

Two other verses which are probably relevant to this topic are:

We have made changes (?) in this quran QurŸån that they might be reminded. [17.41/3]

If we so will, we shall assuredly take away what we have revealed to you. [17.86/8]

In the light of all these verses it cannot be denied that some revision of the quran QurŸån (as it was publicly proclaimed) took place. This was admitted by Muslim scholars in their doctrine of abrogation (nasikh an-nåsikh wa l mansukh wa-l-mansõkh). The idea underlying the doctrine is that certain commands to the Muslims in the quran QurŸån were only of temporary application, and that when circumstances changed they were abrogated or replaced by others. Because the commands were the word of God, however, they continued to be recited as part of the quran QurŸån. Thus the command to spend a considerable part of the night in prayer, given at the beginning of sura 73 was abrogated or cancelled by the long verse [20] at the end, doubtless because in view of the public responsibilities of muhammad Muøammad and the leading Muslims at Medina it was undesirable that they should be awake much of the night. The quotations just given, however, if taken at their face value, indicate something more extensive than is contemplated in the doctrine of abrogation. If due attention is also paid to the words in 75.17 spoken by God (or perhaps the angels) to muhammad Muøammad: 'ours it is to put it together and recite it', the process of 'collecting' separate passages to form suras would also be undertaken by muhammad Muøammad as he followed a divine initiative; the word here translated 'put together', jam jamÿ, is the word later used for the 'collection' of the quran QurŸån after muhammad Muøammad's death. 1

To complete this survey of the possibilities of revision another important passage must be quoted [22.52/If.]:

We have not sent before you (muhammadany Muøammad)any messenger or prophet, but that, when he formed his desire, Satan threw something) into his formulation; so God abrogated what Satan threw in; then God adjusts his signs (or verses) . . . that he may make what Satan has thrown in a test for the diseased of heart and the hard-hearted . . . and that those with knowledge may know that it is the truth from your Lord and believe in it . . .

This verse is usually illustrated by the story of the 'satanic verses' intruded into sura 53 and later cut out 2; but there is nothing in the text of the passage to prevent something similar having happened in a number of other cases. The underlying principle is that something once proclaimed and recited as part of the quran QurŸån came to be regarded as satanic and then was no longer regarded as belonging to the quran QurŸån.

The use of the word 'abrogate' (yansakhu) in this passage differs from its usage in the theory of abrogation, for in the latter the abrogated verses are still retained as part of the quran QurŸån. In passing it may be noted that the retention of abrogated verses in the text of the quran QurŸån as we have it is a confirmation of the accuracy of the text, since it shows that later textual scholars did not remould it in accordance with their own conceptions. The discussion of abrogation in Islam has been voluminous, but belongs primarily to the sphere of jurisprudence. Some of the standard works of the jurists like the 'Epistle' of shafii ash-Shåfiÿi (d. 820), 3 have sections on various questions connected with abrogation, while there are also special treatises on the subject which list and discuss the 'abrogating' and 'abrogated' verses of the quran QurŸån. 4 The fifteenth-century scholar suyuti as-Suyõþæ in his compendium of quranic QurŸånic studies known as the itqan Itqån devotes about half a dozen pages to the question. 5 Many subtle points were raised by the jurists, and the conception was applied not only to the quran QurŸån but also to the Sunna (or practice of muhammad Muøammad), while it was further asserted that the laws of the Jews and Christians had been abrogated by the revelation of the quran QurŸån.

If the later theories of jurists and others are distinguished from what the quran QurŸån itself says, it would seem that various processes took place which may be comprehended under the term 'revision'. It may be conjectured that muhammad Muøammad carried out this revision in accordance with what he understood to be divine guidance. Perhaps this took the form of a repetition of the revelation in the revised form. There is bound to remain some uncertainty about details, but enough has been said to justify an examination of the text of the quran QurŸån to discover detailed evidence of revision.

2. Evidences of revision and alteration

The simplest form of 'revision' is the 'collection' or putting together of the small units in which the revelation originally came. There are grounds for thinking that this process was begun by muhammad Muøammad himself, that is, that it was continuous with his receiving of revelations. This seems to be implied by 75.17 which has already been mentioned. The whole passage runs:

Move not your tongue in it to do it quickly;

ours is the collecting of it and the reciting of it;

when we recite it follow the reciting of it;

thereafter ours is the explaining of it. [75.16-19]

The most likely explanation here of the word 'collecting' (jam jamÿ) is that passages which had originally come to muhammad Muøammad separately were now repeated for him in combination with one another. This explanation is borne out by other points. When muhammad Muøammad's opponents are challenged to produce a sura [10.38/9] or ten suras [11.13/16] like what has been revealed to him, the implication is that there are already in muhammad Muøammad's possession ten units which may be called 'suras'. The date of the second passage is at latest early Medinan, and that would make it possible for many other suras to have been added before muhammad Muøammad's death. Again, it has always been held by Muslim scholars that the mysterious letters are part of the revealed text and were not added by later 'collectors'; and since there is a certain grouping together of the suras with letters (as will be seen in the Table pp. 206-13), it is probable that these groups already existed as groups in muhammad Muøammad's lifetime. If the bismillah bismillåh is also part of the original text, this would be a reason for thinking that the commencement of the suras at least goes back to muhammad Muøammad. Moreover, the great variation in the length of the suras is hardly accounted for by differences of subject, rhyme or form-the type of criterion which might have been used by collectors; and this suggests that much of the quran QurŸån was arranged in suras before the collectors began their work. Altogether, then, it is likely that much of the work of 'collecting' had been performed by muhammad Muøammad guided by a continuing process of revelation.

Next it may be noticed that not only were passages placed together to form suras but that, when this was done, some adaptation took place. One piece of evidence for this is the occurrence of hidden rhymes. 6 It would seem that sometimes, when a passage with one assonance was added to a sura with a different assonance, phrases were added to give it the latter assonance. As an example of this, sura 23.12-16 may be analysed.

(12) la-qad khalaqna khalaqnå l insana l-insåna min sulala sulåla/mm tin þæn

(13) thumma jaalna hu jaÿalnå-hu nutfa nuþfa/fi fæ qararin qarårin makin makæn

(14) thumma khalaqna khalaqnå n nutfata n-nuþfata alaqa ÿalaqa/

fa khalaqna fa-khalaqnå l alaqata l-ÿalaqata mudgha muðgha/

fa khalaqna fa-khalaqnå l mudghata l-muðghata izaman ÿiñåman/

fa kasawna fa-kasawnå l izama l-ÿizåma lahman laøman/

thumma anshana hu anshaŸnå-hu khalqan akhara/

fa tabaraka fa-tabåraka llahu llåhu ahsanu aøsanu l khaliqin l-khåliqin

(15) thumma inna-kum bada baÿda dhalika dhålika la mayyitun la-mayyitõn

(16) thumma inna-kum yawma l qiyamati l-qiyåmati tubathun tubÿathõn

The translation might run as follows:

(12) We have created man of an extract/of clay

(13) Then we made him a drop/in a receptacle sure

(14) Then we created the drop a clot,

then we created the clot a morsel,

then we created the morsel bones,

then we clothed the bones with flesh,

then we produced him a new creature;/

blessed be God the best of creators.

(15) Then after that you are dead,

(16) then on resurrection-day you are raised again.

In this example it is to be noted that the verses as they stand rhyme in i - æ(1 )-in fact in - æn or un - õn-which is the assonance of the sura as a whole. Verse 14, however, is unusually long, and moreover can be broken up into six short verses, five of which rhyme in -a, while the sixth, which is superfluous to the sense, gives the rhyme in in - æn. The same rhyme in -a can also be found in verses 12 and 13 by dropping the concluding phrase. With the omission of the rhyme-phrases verses 12 to 14 constitute a little passage of seven verses rhyming in -a, describing the generation of man as a sign of God's creative power. It may be noted that the word sulala sulåla, translated 'extract' to suit the following phrase, may also mean 'the choicest part of a thing' or 'what is drawn gently out' and so 'semen'; in the only other instance of the word in the quran QurŸån it is stated that, while the first man was created from clay, his progeny came 'from a sulala sulåla of base water' [32.8/7]. Thus the removal of the rhyme-phrases seems to give a better and clearer sense. It may further be supposed that verses 15 and 16 were added as part of the adaptation of the passage to its place in this sura. The passage which immediately follows, 23.17-22, has marks of having been similarly dealt with; when the concluding phrases with the rhymes are detached, there are traces of an assonance in fail fåÿil (namely, taraiq þaråŸiq, fawakih fawåkih, etc.). A number of other passages appear to have been treated in the same way. 7

Of special interest are one or two cases where the rhyme of the sura changes. In sura 3, for example, the first part (up to about verse 20/18) 8 rhymes in a -å(l), and so does the end, from verse 190/87 to verse 200. The large middle section, however, has the rhyme in i - æ(l). Near the point where the first change occurs stands a passage [33/0-41/36] dealing with the story of Mary and Zechariah, in which several of the verses-namely, 37/2, 38/3a, 39a/3b, 40/3 5, 41/36-rhyme in a - å(l), while it seems possible that the other verses have had phrases added to them to carry the rhyme i - æ(l); e.g. the end of 36/1 would be shaytan ash-shayþån if ar rajim ar-rajæm is removed. 9 Thus it looks as if a portion with the rhyme i - æ(1) had been inserted into a sura which originally rhymed in a - å(l) and an attempt made to dovetail the two pieces together. The impression is strengthened when it is noticed that the rhyme i - æ(l) occurs at the end of verse 18/16 carried by a phrase with a difficult construction which leads on to 21/0 rather than to 19/17 and 20/18-19. Other instances of something similar connected with a change of rhyme occur in 13.2-4 and 19.51/2-58/9; but these cases are not so clear.

There are, again, many passages in which the rhyme-phrases can be detached without revealing an older rhyme underneath. In these cases one cannot be certain that revision has taken place, since (as noted above on pp. 70-1) an otiose, and therefore detachable, rhyme-phrase often appears to mark the close of a verse. When, however, such a phrase is found at the end of a number of consecutive verses [as in 6.95-9, 102-4] it is reasonable to assume that it has been inserted into an originally unrhymed passage in order to give it the rhyme of the sura. In two cases [6.84-7; 38.45-8] this seems to have been done with a list of names; and there is something comparable in 19.5 1/2-57/8.

Another way in which passages have been adapted is illustrated by 6.141/2-144/5. These verses cannot be grammatically construed as they stand, but each verse may be divided into two parts. The first parts by themselves give a list of God's bounties in the produce of the soil and animals; but into this list sentences (the second parts) have been introduced combating pagan food-taboos. Again in 7.57/5, 58/6 the sign of God's goodness in the revival of dead land and the varying response of different soils-perhaps a simile of the varying response of men to the divine message-has been transformed by inserted sentences into a corroboration of the resurrection; the insertions are marked by a sudden change of pronoun from 'he' to 'we', referring to God.

In addition to these changes which seem to have taken place when the passage was adapted to its place in a sura, there are many other evidences of revision and alteration. It should be theoretically possible to revise a passage in such a way that no mark of the patching remains, but in practice a careful reader will often be able to detect the alteration through some unevenness in the style. There are indeed many roughnesses of this kind, and these, it is here claimed, are fundamental evidence for revision. Besides the points already noticed-hidden rhymes, and rhyme-phrases not woven into the texture of the passage-there are the following: abrupt changes of rhyme; repetition of the same rhyme-word or rhyme-phrase in adjoining verses; the intrusion of an extraneous subject into a passage otherwise homogeneous; a differing treatment of the same subject in neighbouring verses, often with repetition of words and phrases; breaks in grammatical construction which raise difficulties in exegesis; abrupt changes in the length of verses; sudden changes of the dramatic situation, with changes of pronoun from singular to plural, from second to third person, and so on; the juxtaposition of apparently contrary statements; the juxtaposition of passages of different date, with the intrusion of late phrases into early verses. So common are these features in the quran QurŸån that they have often been regarded as characteristics of its style and in no need of further study or explanation. This is not the case, however. It is here being argued that these features of the quran QurŸån are most simply explained by supposing a measure of revision and alteration; but even if this view is rejected, some explanation of these features is still called for. Meanwhile what has been said about the unevenness and roughness of quranic QurŸånic style may be amplified.

Glosses are a common feature of ancient Greek, Latin and other manuscripts. They are short explanations of some obscurity, presumably first written on the margin by some reader and then mistakenly incorporated in the text by a later copyist. While it is doubtful if the quran QurŸån contains any glosses in the strict sense, there is something approaching a gloss in 2.85/79. Beginning at the previous verse the passage runs:

(Recall) when we made a covenant with you (on the following terms): You shall not shed your own (sc. one another's) blood; and you shall not expel yourselves from your dwellings. Then you confirmed it, yourselves being witnesses.

            Then there you are killing yourselves, and expelling a party of you from their dwellings, as you join together against them in guilt and enmity;/and if they come to you as prisoners, you shall ransom them;/and it is forbidden to you, their expulsion. Do you believe . . .

The clause about ransoming prisoners seems an intrusion here. Bell in his Translation considers that it belongs to the terms of the covenant in the previous verse, which is possible but not certain. 10 If this clause is removed, the following clause, which may then be translated 'although it is forbidden to you' is perfectly clear without the addition of 'their expulsion', ikhraju hum ikhråju-hum. There is thus a strong presumption that 'their expulsion' is a gloss or addition, made after the clause about ransoming prisoners had been intruded. Other possible examples of such additions or explanatory substitutions will be found in: 6.12, 20; 7.92/0; 21.47/8, 104; 27.7; 41.17/16; 76.16. 11

Explanations of unusual words or phrases are sometimes added in the form of an extension of the passage. There are twelve instances of such extensions beginning with the words: 'What has let you know what . . . is?' 12 A short description then follows. It is clear that some of the descriptions have been added at a later time, since they do not correspond to the sense in which the word or phrase was originally taken. The most striking case is at the end of sura 101 [verses 9/6-11/8]: ....'his mother shall be hawiya håwiya. And what has let you know what it is? A scorching fire.' hawiya Håwiya presumably meant 'childless' owing to the death or misfortune of her son; but the addition suggests that it is a name of Hell. A somewhat similar passage is 90.12-16. The addition is seldom an exact definition of what is to be explained.

Additions and insertions of other kinds may be illustrated from the shorter suras. In sura 91 it is evident that the main passage, when first revealed, ended at verse 10; but this is followed by a summary of the story of thamud Thamõd, which may either have been added to illustrate the moral, or simply placed here because of the similar rhyme. Verses 6 and 7 of sura 88 may be marked as an insertion by the different rhyme, and verses 33 and 34 of sura 78 by the breaking of the connection between verses 32 and 35. In sura 87 a sudden change in the dramatic situation at verse 16 marks an addition which might have followed immediately on the original revelation, but is probably much later. In sura 74 the passage 31-31/4 is clearly marked as an insertion by the different style and length of verse. Some of these additions might conceivably be due to a later collector or reader; but this is unlikely.

There are other additions, however, which can hardly have been made without authority. The misplaced phrase of 2.85/79, for instance, though it looks like a gloss written on the margin and taken in by a copyist at the wrong place, makes a real addition to the regulation laid down. There are few such misplacements, but short additions which make substantial alterations to the sense are frequent. In 74.56/5 we have a limitation of the freedom of man's choice which virtually takes back what had been stated in verse 55/4; cf. 76.29, 30; 1.28, 29. This corresponds to the hardening of the doctrine of predestination which took place in Medinan days. Reservations introduced by illa illå, 'except', are specially frequent. We must not, of course, assume that every such reservation is a later addition, but in a number of cases there are independent reasons for such an assumption, as in 87.7, and 95.6, where illa illå introduces a longer verse, which has characteristic Medinan phraseology, into an early passage with short rhythmic verses. Such additions, making as they do a distinct modification of the statement, must have been deliberately introduced. In at least some of them we can discern the grounds for making the exception.

Longer additions can sometimes be easily distinguished. Thus in sura 73 a long verse occurs at the end which, by containing a reference to Muslims engaged in fighting, is clearly marked as Medinan, and is recognized by everyone as being so. But the rest of the sura, and especially the beginning, is in the short crisp verses characteristic of early passages. The reason for the addition is that the passage at the beginning recommended night-prayer; but since this was being overdone, it became necessary in Medina to counsel moderation. 13

Additions in the middle of suras are common. A few examples will suffice. The first part of sura 19 has the assonance in iyya - iyyå, but this is interrupted by verses 34/5 to 40/1, which have the common i - æ(I) assonance. These verses follow an account of Mary and Jesus, and, by rejecting the idea of God having offspring, criticize a popular misconception of Christian doctrine. 3.130/25-134/28 warn against the taking of excessive interest, and promise heavenly reward to those who act generously. The passage evidently closed with the rhyme-phrase of 134/28, but two verses follow giving a further description of those who do well by repenting and asking forgiveness, and containing a promise of heavenly reward which is largely a repetition of that already made. Those who have transgressed but are prepared to reform are thus included. Verses 5 to 8 of sura 22 argue for the resurrection as in line with God's power otherwise manifest, and close by scoffing at those who 'without knowledge, guidance, or light-giving book' argue to the contrary. Verses 9, 10 join to this rather awkwardly and threaten not only future punishment but 'humiliation in this life', a Medinan threat, to those who so act. The change of tone and attitude shows clearly enough that these verses did not belong to the original passage. In sura 37.73-132 there are accounts of various Biblical persons, closing in three cases with the refrain: 'Thus do we reward those who do well; verily he is one of our servants believing'. In the case of Abraham, however, this refrain [110f.] is followed by a statement about the posterity of Abraham and Isaac [112f.]. This must have been added after the passage was composed.

Another important feature of quranic QurŸånic style is that in many cases a passage has alternative continuations, which follow one another in the present text. The second of the alternatives is marked by a break in sense and by a break in grammatical construction, since the connection is not with what immediately precedes, but with what stands some distance back; there may also be the repetition of a word or phrase. Thus 23.63/5, which speaks of men continuing a defective course of conduct, is followed by three passages introduced by hatta øattå idha idhå, 'until when', commencing with 64/6, 77/9 and 99/101 respectively. It is possible, with some straining, to join verse 77/9 to 76/8, but verse 99/101 will not join to 98/100. The words hatta øattå idha idhå, however, require before them a reference to something continuing. Verses 99/101f. are in fact the proper continuation of 63/5, as is evident if we read them together; the other verses introduced by hatta øatta idha idhå are alternative and presumably later continuations of 63/5. Again, 5.42/6 begins with a phrase samma sammåÿ õna li-l-hadhib, which is entirely out of connection with what precedes. The same phrase occurs in 41/5, however; and, if the part of 41/5 from this phrase to the end is omitted, 42/6 fits perfectly to the earlier part of 41/5. Here also then, there are alternative continuations. Another example will be found at the end of sura 39 where there is a verse which appears isolated [75]. It follows a Judgement scene and evidently belongs to it; but the scene is already finished; judgement has been given, the unbelievers have been sent to Gehenna, the pious have entered the Garden; then we find ourselves back at the scene of Judgement where judgement will be given with truth. This phrase, which has already occurred in verse 69, indicates what was the original position of verse 75; it followed the first phrase of 69 and completed the scene; at some later stage it was displaced by the much longer description in verses 69-74. 14

Occasionally a change of rhyme may accompany such a substitution. Thus 80.34-37 have their assonance in ih - æh, while verses 38-42, which join equally well to verse 33, have the -a assonance which runs through the whole of the rest of the sura. More frequently the occurrence of the same rhyme-word or -phrase is a sign that such a substitution has been made, since the new version ends with the same rhyme as that which it replaced. Thus in sura 2 verses 102/96 and 103/97 both end in law kanu kånõ yalamuna yaÿlamõna, 'if they had known', which gives a presumption that the latter verse was intended to replace the former. In sura 3 the similar ending indicates that verse 144/38 is a substitute for verse 145/39. A similar phenomenon is to be found in 9.117/8, 118/9; 34.52/I, 53/2; 45.28/7, 29/8; 72.24/5, 27-8. In such cases the alternative continuations often stand in reverse order of date, the later coming first, but this is not an invariable rule. 15

Further evidence of alteration and revision may be obtained by approaching the quran QurŸån from the standpoint of the subject-matter and considering passages dealing with the situations which presented special difficulties or problems to muhammad Muøammad and the Muslims. In these passages there is often much confusion. A simple case is that of the ordinance concerning fasting. When he removed to Medina, muhammad Muøammad hoped for support from the Jews and showed himself willing to learn from them. Tradition says that he introduced the Jewish fast of the ashura ÿAshõrå, which consisted of the Day of Atonement preceded by some days of special devotion. Later, the month of ramadan Ramaðån was prescribed. Now, in 2.183/79-185/1 these two things lie side by side; verse 184/0 prescribes a fast of a certain number of days, verse 185/1 the month of ramadan Ramaðån. The two verses are, of course, read consecutively, and the 'certain number of days' of the former verse is held to be made more precise by the mention of the month of ramadan Ramaðån in the latter. But 'a certain number of days' is not naturally equivalent to a month, and the repetition of phrases in the two verses shows that the one was intended to replace the other. The verses are, in fact, alternative continuations of 183/79. 16

The marriage laws in sura 4 are another clear case of alternative continuations. Verse 23/7 lays down the forbidden degrees of relationship, and reproduces the Mosaic list with some adaptation to Arab custom. That this was deliberate is shown by 26/31, which states that 'God desires . . . to guide you in the customs of those who were before you'. At a later time, however, some relaxation became necessary, and 25/29-30 and perhaps 27/32a were substituted for 26/31, allowing marriage with slaves. Finally 24/8, which gives ample liberty, was substituted for 25/29-30, and 28/32b was added to give a verse-ending. The similar endings of 26/31, 27/32a and 28/32b show that substitutions have been made.

The change of qibla (the direction to be faced in prayer) affords another example. The passage dealing with it is very confused [2.142/36-152/47]; the portion from 141/39 especially is unintelligible as it stands. When analysed, however, the verses turn out to contain (a) a private revelation to the Prophet of the solution to his problem [144/39a, 149/4]; (b) a public announcement, using part of (a) accompanied by an appeal for obedience based on gratitude [144/39a, 150/45-152/47]; and (c) the final form of the ordinance [I44/39a, 144/39b].

The process of the introduction of the religion of Abraham is outlined for us in 2.130/24-141/35. It takes the form of answers to the assertion of Jews and Christians [135/29a]: 'They say: "Be ye Jews or Christians and ye will be guided".  This is followed by three retorts introduced by 'Say'. Verses 139/3-141/35 claim that the Prophet and his followers have a perfect right to serve God in their own way, as did Abraham and the patriarchs; these constituted an independent religious community long since passed away. This passage was cut off and replaced by 136/0, 138/2, in which it is claimed that muhammad Muøammad and his followers stand in the line of Abraham and the patriarchs, Moses, Jesus and all the prophets. It was again modified by the insertion of 137/1 in place of 138/2. Finally, the short retort of 135/29b was written in, professing the creed of Abraham, who was a hanif øanæf and no polytheist. Verses 130/24 to 134/28 are a further addition.

The question of the pilgrimage, which was part of the religion of Abraham, also caused difficulty. The ceremony was recognized and muhammad Muøammad's followers were counselled to take part in it, but as hanif øanæf, followers of the religion of Abraham, not as polytheists [22.31/2]. Sacrificial animals were to be sent to Mecca [22.34/5a, 33/4]. When, however, Muslim attacks on Meccan caravans, and especially the battle of Badr, led to bloodshed, it became dangerous for any Muslim to visit Mecca. It was therefore laid down that the animals dedicated for sacrifice might be slaughtered at home and their flesh given to the poor. This can be deduced from 22.29/30-37/8. 17

Fighting in the sacred months also caused difficulty. muhammad Muøammad's attitude is made clear by the analysis of sura 9. These months were at first recognized as a period of truce, by a deliverance which consisted of 9.3 6a, 2, 5; but since the intercalary month, which kept the Arab lunar year in conformity with the seasons, was decreed from Mecca, misunderstandings about which months were sacred would soon arise. Hence the deliverance was issued which now stands as 9.36,37, abolishing the intercalary month and decreeing that war with the polytheists was to be carried on continuously. The discarded verses dealing with the sacred months now appear as verses 2 and 5, linked up with a renunciation of agreements with polytheists, probably the treaty of hudaybiya al-Øudaybiya. As the heading informs us, however, this is also a proclamation to be made at the pilgrimage; and it was presumably altered and added to for this purpose after the fall of Mecca. 18

The defeat of the Muslims at uhud Uøud was a severe blow to the confidence of the Muslims. The passage dealing with the battle is in great confusion [3.102/97-179/4]. Analysis shows that there was an address intended for delivery before the battle, which consisted of verses 102/97, 103/98-9, 112/06a, 115/1-117/3, 123/19, 139/3-143/37, 145/39-151/44, 158/2, 160/54. Part of this, perhaps from 139/3 onward, was re-delivered, with a few alterations, some time after the battle. Reactions to the defeat appear in a reproof to the Prophet himself for having, without authority, promised the assistance of angels [verses 121/17, 124/0, 125/1 and parts of 126/2-129/4]. That was later revised as an explanation and rebuke to his followers. That he had been inclined to speak angrily to them is indicated in the private verse [159/3]. Part of this 'rough' speech may be embedded in 152/45-154/48, a passage which has been revised and added to in a milder sense later. In fact, we can see the attitude to the defeat growing gradually calmer and more kindly towards the faithful. Finally, when the set-back had been overcome, part of the original address was used again, with a new continuation added after 110/06a, in preparation probably for the attack on the Jewish tribe of nadir an-Naðær [110/06b-114/0]. 19

The great volume of evidence, of which what has been presented here is only a sample, shows that the quran QurŸån is far from being a straightforward collection (out of chronological order) of short passages of a revealed text. The matter is too complex for any simple explanation of this kind. The vast number of dislocations and the roughness of some of them cannot simply be ascribed to 'the quranic QurŸånic style'. The modern scholar may seldom be able to give a correct solution of the problems raised by the dislocations, but it can surely be no longer denied that there are problems of this kind. My personal view is that in the working out of solutions to these problems in his Translation Richard Bell was often successful. The following section, however, presents a small part of his view which has not met with the same degree of approval.

3. Bell's hypothesis of written documents

(a)        The hypothesis.

The critical literary analysis of the quran QurŸån, besides producing the kind of evidence given in the last section led Richard Bell to formulate a particular theory about the place of written documents in the 'collection' of the quran QurŸån. This theory was not simply that parts of the quran QurŸån had been written down at a fairly early stage in muhammad Muøammad's career, but more particularly that the occurrence in the middle of a sura of a passage wholly unrelated to the context was to be explained by the supposition that this passage had been written on the back of the 'scrap of paper' used for one of the neighbouring passages which properly belonged to the sura. Bell used the word 'paper' as a convenient term for any kind of writing material.

As examples of such passages out of relation to the context Bell selected 75.16-19, 84.16-19, and 88.17-20. The argument may be presented most clearly in the case of the latter. The sura begins with a description of the Judgement and the fate of the wicked, and then continues with a picture of the righteous

(10) in a garden lofty (aliya ÿåliya)

(11) wherein they hear no babbling; (laghiya låghiya)

(12) therein is a spring running; (jariya jåriya)

(13) therein are couches upraised (marfua marfõÿa )

(14) and goblets set out (mawdua mawdõÿa)

(15) and cushions in rows (masfufa maãfõfa)

(16) and carpets spread. (mabthutha mabthõtha)

(17) Will they not look at the camels, how they have been created (khuliqat)

(18) at the heaven, how it has been uplifted; (rufiat rufiÿat)

(19) at the mountains, how they have been set up; nusibat nuãibat)

(20) at the earth, how it has been laid flat? (sutihat suþiøat)

(21) So warn. You are only a warner . . . (mudhakkir)

The argument here is as follows. The passage 17-20 has no connection of thought either with what goes before or with what comes after; and it is marked off by its rhyme. It is thus difficult to know why it has been placed here. If one assumes that its position has been given to it by a collector, one may still ask whether a responsible collector could not have found a more suitable place for it. Bell's hypothesis is that verses 17-20 have been placed here because they were found written on the back of verses 13-16. He further holds in this particular case that 13-16, which are marked off by rhyme from the preceding verses, were a later addition to these, and happened to have been written on the back of a 'scrap' which already contained 17-20.

Something similar is true of 75.16-19. Verses 13-16 (partly distinguished by rhyme, partly by length) seem to have been added to 7-12, which deal with the Last Day, and to have been written on the back of the early 'scrap' containing 16-19 (quoted above on p. 89). In sura 84 there is no abrupt change of rhyme, but verses 13-15 destroy the balance of the preceding piece, verses 7-12, which is complete as it stands. In each case, then, an addition has been made, and the addition occupies approximately the same space as the extraneous passage which follows. A simple explanation of the position of the extraneous passage would thus be to suppose that it stood on the scrap of paper on which the addition was later written, and that the two sides of the paper had been read and copied consecutively when the quran QurŸån came to be made up in the form of a codex.

Similar examples may be found throughout the quran QurŸån. To take an example from near the beginning: 2.15/16 compares those who have accepted the Prophet's guidance and then gone back upon it to people who have lit a fire, which has then gone out, leaving them blinded in the darkness. Verse 18/17, 'Deaf, dumb and blind, they do not return', evidently closes the passage, but verses 19/18, 20/19 contain another simile: they are like people in a thunder-storm, the rain pours down, the thunder deafens them, the lightning blinds them. Evidently this is a parallel to 17/16 and should have preceded 18/17. It has been added later. There follows a passage, 21/19b, 22/20, unconnected with the context, appealing for the worship of God and adducing signs of his power and bounty. This appears to be continued, after a break, in 28/26, 29/27. Now 27/25, while not evidently an addition, is probably so, for 26/24 finishes with a reference to the 'reprobate', which is conclusive enough. But 27/25 proceeds to describe a special class of 'reprobates', who violate a covenant after having made it. Further, we find in verses 163/58-165/0a a passage which, by the use of the rather unusual word andad andåd,  'peers' is marked as almost certainly a continuation of 21/19b, 22/20, 28/26, 29/27. Here we have, not preceding but following, a passage 165/0b-167/2, which returns to the theme of 161/56, 162/57, and must have been intended as an addition to that passage. This whole section is an interesting example of how a passage has been expanded by additions. The point, however, here is that we find a passage originally dealing with the worship of God apparently cut up, and the back of the pieces used for making insertions into other passages.

An interesting example of the same kind is found in sura 9. The last two verses of this sura are traditionally said to have come to the knowledge of Zayd thabit ibn-Thåbit when he had almost completed his task of collecting the quran QurŸån, and were placed here as the most convenient position at the time. This is evidently an attempt to account for the fact that there is a break in connection between verses 127/8 and 128/9, and another between 128/9 and 129/30. These two verses seem to stand isolated, but 129/30 will connect well enough with 127/8, though the latter verse ends as if nothing more were to be said. It is a case of something having been later added to a passage, and we may suppose that the back of 128/9 was used to write it on. By some accident (127/8 had itself been used for the writing of another passage) the back was read by the compilers before the addition. But this is not all; verse 40 of the same sura stands isolated, though it evidently requires something in front of it. The pronoun 'him' must evidently refer to the Prophet of whom there has been no mention in the context, but verse 128/9 speaks of the Prophet, and if we read verses 128/9 and 40 together we get a moving appeal for loyalty to the Prophet addressed to his followers. This has evidently been cut in two, one part being added to 127/8 and the other placed after 39.

The reverse seems also to have taken place; scraps of paper were somehow pasted together to form a sheet. Sura 14.8-14/17-an evident addition to the account of Moses-in which he addresses his people in regular quranic QurŸånic style, is followed by a series of disjointed pieces, 15/18-17/20, 18/21, 19/22, 21/4-22/7, 23/8, which together occupy practically the same space. In fact, it is almost a rule in the later parts of the quran QurŸån that an addition or connected deliverance of any length is preceded or followed by a number of disconnected pieces which together make up approximately the same length. An interesting instance of this occurs at the end of sura 2. There we find a long deliverance dealing with the recording of debts [282, 283]. This occupies approximately the same space as verses 278-281, a deliverance forbidding usury, 284 a separate verse, and 285, 286 a profession of faith of the believers. Into this piece two little sentences intrude at the junction of the verses [285b, 286a]; they have no connection with each other or with the context and break the connection of v.285 and v. 286, which must have originally formed one verse. If now we suppose the deliverance regarding debts [282-3], to have been written on the back of a sheet (or part of a sheet) which contained the deliverance on usury [278-81], and on that of a second sheet containing 284-6, we find that the intrusion into the latter piece comes practically opposite a proviso introduced into the deliverance about debts excepting from its scope transactions in the market where goods pass from hand to hand. This proviso, we may suppose, was written on the back of two scraps and inserted into the deliverance. To do so, the sheet was cut and the proviso pasted in. In this way the two extraneous scraps appear on the other side of the sheet.

The same thing occurs in sura 4, where, if we suppose verses 88/90-91/3 to have been written on the back of 79/81-87/9, a proviso introduced by illa illå [90/2a], will come opposite 82/4 which breaks the connection between 81/3 and 83/5. This part of the sura is further interesting in that the passage 79/81-81/3, 83/5, 84/6 is almost certainly private and was not meant to be publicly recited. A number of private passages of this kind, intended only for muhammad Muøammad himself, are included in the quran QurŸån. The most striking of them is 3.159/3, which can hardly have been intended for publication either at the time or later; cf. also 154/48c and 161/55. The passage about fasting discussed above (p. 98) gives a further illustration [2.183/79-187/3]. Verse 186/2 is entirely unconnected; it has no reference to fasting, and while in the preceding verses the believers are being addressed and God is spoken of in the third person, in this verse God is speaking, the Prophet is being addressed, and other men are spoken of in the third person. Verse 187/3 returns to the subject of fasting and the dramatic setting of 183/79-186/2. If we consider the length of 185/1, we shall find that when written out it occupies approximately the same space as 184/0 and 186/2 together. The presence of this latter verse seems to have arisen from the necessity of adding to the space afforded by the back of 184/0 by using the back of a verse from some other context.

(b) Critique of the hypothesis.

This theory that the order of the quran QurŸån is often due to the fact that some passages were written on the backs of others was worked out in detail for the whole quran QurŸån by Richard Bell. His results are incorporated in his Translation by various typographical devices such as divisions down the middle of pages. The more one studies these results, the more one is impressed by the infinite pains taken and the great ingenuity shown. For a long time to come scholars will have to take account of this detailed work.

The hypothesis certainly cannot be rejected out of hand. For one thing it seems clear that there were written documents from a fairly early period. Even if muhammad Muøammad himself did not write, he could have had them written by secretaries. It is known that he used secretaries in his later years, and there are Traditions in which the secretaries are employed to write down the revelation. The reference to muhammad Muøammad's forgetting in 87.6 could be held to suggest the inference that he came to distrust his memory and wrote out and memorized the revealed messages before proclaiming them publicly. The gibe of the Meccans about 'old-world tales which he has had (?) written for himself' implies that at Mecca he was at least suspected of having things written down [25.6]. If, as is likely, he had some of the quran QurŸån written, he may have tried to keep the matter secret. At Medina one would expect that at least the legal deliverances were recorded. The report about the first 'collection' of the quran QurŸån after muhammad Muøammad's death by Zayd thabit ibn-Thåbit implies that some was already written on pieces of papyrus and other materials. The result of Zayd's work was a 'collection' of the quran QurŸån on 'sheets' (suhuf ãuøuf), and these eventually passed into the possession of hafsa Øafãa. As was argued above, it is unlikely that there was any official 'collection' such as is described; but it is fairly certain that hafsa Øafãa had 'sheets' of some sort. It is thus probable that much of the quran QurŸån had been written down in some form during muhammad Muøammad's lifetime. It is even possible that there were several written versions of parts of it in the hands of different individuals.

Bell's distinctive hypothesis, however, is concerned not with the mere existence of written documents, but with a special way of dealing with them which he alleges to be responsible for some aspects of the order of the text. It should at once be admitted that what he suggests may occasionally have happened. On the other hand, there are suras (such as 80 and 96) where unconnected pieces have been brought together; and Bell apparently simply accepts this fact without trying to apply his theory. It may then be inferred that, at least at some periods, whoever was responsible for collecting the quran QurŸån was not unduly worried by the absence of continuity of thought; and in so far as this is the case a discontinuity of thought in a sura may easily have come about without the passage having been written on the back of another. This makes some of Bell's elaborate reconstructions (such as the examples from sura 2 and sura 9) all the more dubious.

It may also be urged that little is gained by the hypothesis. The problem before the scholar is the accidental character of the unconnected passages. In effect Bell's hypothesis explains this accidental character by supposing another accident, namely, that one passage was written on the back of another. In particular cases there is bound to be great uncertainty about the precise way in which the hypothesis is to be applied; but, even if the application were known to be correct, little would be added to our understanding of early Islam. In this respect the results produced by the hypothesis are in contrast to the evidence tending to show revision and alteration. If the analysis of the passage about the qibla is sound, then it gives us increased insight into the profound re-orientation of the policy of the Islamic state about March 624.

The emphasis on documents in the hypothesis and in Bell's treatment generally requires to be balanced by giving increased weight to the aspect of oral transmission. In the traditional account 'the hearts of men', that is, their memories, was one of the sources drawn on by Zayd thabit ibn-Thåbit; and the quran QurŸån reciters subsequently became an important group of men. The possibility that muhammad Muøammad might forget a passage envisaged in 87.6 implies that for at least a time he was relying on his memory. This suggests the further question whether muhammad Muøammad clearly distinguished between proclaiming from memory a message he had previously received and proclaiming at the moment of revelation a message which partly coincided with another message previously received. The reference to God's 'collecting' of the quran QurŸån in 75.17 would seem to imply that muhammad Muøammad received revelations combining (and perhaps adapting) previous revelations. This further implies that a revelation may be repeated, perhaps in slightly different terms. This becomes all the more significant when one remembers the numerous repetitions of phrases and verses throughout the quran QurŸån. It may also be linked up with the phenomenon of alternative continuations. It seems likely, then, assuming that some passages had been revealed in slightly different forms on different occasions, and remembered by individual Muslims in their different forms, that the 'collectors' had on their hands a formidable problem. They would not want to omit any smallest scrap of genuine revelation, and yet the total mass of material may have been so vast that they could not include it all. This may explain some of the roughnesses in the uthmanic ÿUthmånic text.

In conclusion one may underline the value of detailed studies of the text of the quran QurŸån such as those carried out by Richard Bell. At the same time one may urge on scholars the need for concentrating on those aspects of the subject which are likely to contribute to a deepening understanding of the early life of the Islamic community.

quran